Shanulhaq's picture
Update app.py
1447393 verified
raw
history blame
7.55 kB
import streamlit as st
import google.generativeai as genai
import os
from groq import Groq, GroqError
import whisper
from gtts import gTTS
import tempfile
import logging
import numpy as np
from pydub import AudioSegment
# Set up logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
# Securely configure Google API Key for generative AI (set this in the deployment environment)
GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
genai.configure(api_key=GOOGLE_API_KEY)
# Set up Groq API key for audio processing
groq_api_key = os.getenv('groq_api_key')
if not groq_api_key:
raise ValueError("GROQ_API_KEY is not set.")
try:
groq_client = Groq(api_key=groq_api_key)
logger.info("Groq API key is set and client is initialized.")
except GroqError as e:
logger.error(f"Failed to initialize Groq client: {e}")
raise
try:
# Load Whisper model for audio transcription
whisper_model = whisper.load_model("base")
logger.info("Whisper model loaded successfully.")
except Exception as e:
logger.error(f"Failed to load Whisper model: {e}")
raise
# Initialize the Generative Model for heart health chatbot
model = genai.GenerativeModel(
'gemini-1.5-flash',
system_instruction=(
"Persona: You are Dr. Assad Siddiqui, a heart specialist. Only provide information related to heart health, symptoms, and advice. "
"Ask users about their heart-related symptoms and provide consultation and guidance based on their input. "
"Always provide brief answers. If the inquiry is not related to heart health, politely say that you can only provide heart-related information. "
"Responses should be in Urdu written in English and in English."
)
)
# Function to get response from the Generative AI chatbot
def get_chatbot_response(user_input):
response = model.generate_content(user_input)
return response.text.strip()
# Function to process audio using Whisper, Groq, and gTTS
def process_audio(audio_file):
try:
# Transcribe audio using Whisper
result = whisper_model.transcribe(audio_file)
user_text = result['text']
logger.info(f"Transcription successful: {user_text}")
except Exception as e:
logger.error(f"Error in transcribing audio: {e}")
return "Error in transcribing audio.", None
try:
# Generate response using Groq API
chat_completion = groq_client.chat.completions.create(
messages=[
{"role": "user", "content": user_text}
],
model="llama3-8b-8192",
)
response_text = chat_completion.choices[0].message.content
logger.info(f"Received response from Groq API: {response_text}")
except GroqError as e:
logger.error(f"Error in generating response with Groq API: {e}")
return "Error in generating response with Groq API.", None
try:
# Convert response text to speech using gTTS
tts = gTTS(text=response_text, lang='en')
audio_file = tempfile.NamedTemporaryFile(delete=False, suffix='.mp3')
tts.save(audio_file.name)
logger.info("Text-to-speech conversion successful.")
except Exception as e:
logger.error(f"Error in text-to-speech conversion: {e}")
return "Error in text-to-speech conversion.", None
return response_text, audio_file.name
# Streamlit page configuration
st.set_page_config(
page_title="Heart Health & Audio Chatbot",
page_icon="πŸ‘¨β€βš•οΈ",
layout="centered",
initial_sidebar_state="collapsed",
)
# Background image and custom CSS
st.markdown("""
<style>
.stApp {
background-image: url('https://cdn.wallpapersafari.com/29/34/8Ak1Sf.png');
background-size: cover;
background-position: center;
background-attachment: fixed;
}
.chat-bubble {
background-color: #128c7E;
color: white;
padding: 10px;
border-radius: 10px;
max-width: 70%;
}
.user-bubble {
background-color: #075e54;
color: white;
padding: 10px;
border-radius: 10px;
max-width: 70%;
align-self: flex-end;
}
img.avatar {
width: 50px;
height: 50px;
border-radius: 50%;
}
</style>
""", unsafe_allow_html=True)
# Custom header
def load_header():
st.markdown("""
<div style="padding:10px;text-align:center;color:white;">
<h1>Heart Health & Audio Chatbot πŸ«€</h1>
<p>Ask me anything about heart diseases or process audio queries!</p>
</div>
""", unsafe_allow_html=True)
# Initialize session state for chat history
if "history" not in st.session_state:
st.session_state.history = []
# Avatar URLs
user_avatar_url = "https://img.freepik.com/free-photo/sad-cartoon-anatomical-heart_23-2149767987.jpg"
bot_avatar_url = "https://img.freepik.com/premium-photo/3d-render-man-doctor-avatar-round-sticker-with-cartoon-character-face-user-id-thumbnail-modern-69_1181551-3160.jpg"
# Function to display chat history
def display_chat_history():
for chat in st.session_state.history:
if chat["role"] == "user":
st.markdown(f"""
<div style="display: flex; justify-content: flex-end; align-items: center; margin-bottom: 10px;">
<div class="user-bubble">
<p><b>You:</b> {chat['content']}</p>
</div>
<img src="{user_avatar_url}" class="avatar"/>
</div>
""", unsafe_allow_html=True)
else:
st.markdown(f"""
<div style="display: flex; align-items: center; margin-bottom: 10px;">
<img src="{bot_avatar_url}" class="avatar">
<div class="chat-bubble">
<p><b>Bot:</b> {chat['content']}</p>
</div>
</div>
""", unsafe_allow_html=True)
# Main application layout
def main():
load_header()
with st.container():
display_chat_history()
# User input for text-based chat
with st.form(key="user_input_form", clear_on_submit=True):
user_input = st.text_input(
label="Type your message...",
placeholder="Ask about heart health...",
max_chars=500
)
submit_button = st.form_submit_button(label="Send")
if submit_button and user_input.strip():
with st.spinner("Thinking..."):
bot_response = get_chatbot_response(user_input)
# Update chat history
st.session_state.history.append({"role": "user", "content": user_input})
st.session_state.history.append({"role": "bot", "content": bot_response})
# Refresh chat display
display_chat_history()
# File uploader for audio processing
st.subheader("Upload audio for processing")
audio_file = st.file_uploader("Choose an audio file", type=["mp3", "wav"])
if audio_file is not None:
with st.spinner("Processing audio..."):
response_text, audio_response = process_audio(audio_file)
st.text(f"Response: {response_text}")
st.audio(audio_response)
# Run the app
if __name__ == "__main__":
main()